← Hub
中文 EN
// Cross-Site Synthesis

Cross-Site Synthesis

Turning an archive into a network, instead of just a longer list
What this is: An ordinary topic site answers "what is X." A synthesis essay answers "why X is the same thing across several of my sites — and where it stops being the same." When one structure keeps turning up in different fields, a result you worked hard for in one of them can simply be collected in another; this line goes looking for those channels. Every source is a page already published on this hub, and every section links straight back to it.
Syn 1 · Four fields, one number Entropy: The Same Number in Five Rooms Physics, information theory, machine learning and bioenergetics compute the same entropy — not a similar-looking formula, the same number up to a conversion constant. All four ask one question: given what I can see, how many possibilities remain underneath? Landauer's principle makes bits and joules convertible, and settled Maxwell's demon after seventy years. The fifth room, which borrowed only the shape of the word, comes last. 10 source pages · physics · mathematics · ai-ml · meta-knowledge · deep-reading · mental-models Read the essay → Syn 2 · One hand, three civilizations Evolutionary Games: One Hand, Three Civilizations Evolution dealt every human society the same hand — kin altruism, reciprocal altruism, tit-for-tat, group boundaries. What differs is which card each environment pushed to the front. The differential order, the spirit of contract, caste-and-renunciation are three stable equilibria natural selection sifted out on three different boards; and philosophy is the manual each writes for an operating system already running. Knowing it's the same hand is what lets you learn from how the others play it. 9 source pages · deep-reading · mental-models · psychology · philosophy · thinker-arena Read the essay → Syn 3 · Three ledgers, one flipped table Cooperation: Three Ledgers, and One Flipped Table "Why does cooperation exist?" Evolutionary biology bills it to the genes (rB>C), game theory to repetition (tit-for-tat), international relations to credible commitment — three ledgers that dissolve altruism into self-interest, all sitting on the same Prisoner's Dilemma, all breaking it with one shape: turning the short-run gain from defecting into a cost across the future. Buddhism's bodhisattva path refuses to count; it dissolves the "self" that would be the beneficiary. Ends with a checklist you can tick line by line, and the naturalistic fallacy. 15 source pages · mental-models · meta-knowledge · deep-reading · civics-geopolitics · buddhism · philosophy · psychology · leadership · sales Read the essay → Syn 4 · A rubber hand and a 2,500-year-old argument The Self: A Rubber Hand and a 2,500-Year-Old Argument The Buddhist mastery argument — a real self would answer to you, and body and mind refuse to — and the rubber-hand illusion are the same inference run twice: one from everyday observation, one from controlled intervention. Two roads with no methodological overlap perform the same operation on "I": moving it out of the subject position. The load-bearing evidence is decomposability — every one of the self's five layers can be peeled loose in a lab. The sharpest boundary: the same silence was reported as opposite conclusions by Vimalakīrti and by Śaṅkara. 15 source pages · buddhism · neuroscience · philosophy · psychology · deep-reading · book-recommendations Read the essay → Syn 5 · A metaphor that was wrong Attention: A Borrowed Metaphor That Was Wrong, and the Nine BLEU It Bought In 2014 an intern bolted a module onto a translation model and called it "attention" — while the neuroscience site says flatly that what the brain does and what the machine does are strangers in mechanism. The analogy was wrong; the invention was right anyway. What crossed the border was not a mechanism but the shape of a constraint: a channel with a hard ceiling, a far larger candidate set, an allocation decision due now. The load-bearing evidence is those nine BLEU; the sharpest boundary is hard gating versus soft weighting, which decides whether a conclusion from one side can move to the other. 12 source pages · neuroscience · ai-ml · cs-papers-deepread · psychology · leadership · mental-models · biographies · deep-reading · book-recommendations · parenting Read the essay →

Coming Up

Syn 6 · Memory — Is "model memory" the same thing as human memory? An LLM's context window isn't memory, and neither is what sits in the weights — a confusion that costs real money in product design.

Further Out

One essay a week, working down the roadmap in order. The full list (17 essays) lives in TOPICS.md.